Postconditioning reduces infarct size via adenosine receptor activation by endogenous adenosine

Transcription

1 Cardiovascular Research 67 (2005) Postconditioning reduces infarct size via adenosine receptor activation by endogenous adenosine Hajime Kin a, Amanda J. Zatta a, T, Mark T. Lofye a, Bradley S. Amerson a, Michael E. Halkos a, Faraz Kerendi a, Zhi-Qing Zhao a, Robert A. Guyton a, John P. Headrick b, Jakob Vinten-Johansen a a Cardiothoracic Research Laboratory, Department of Cardiothoracic Surgery, Carlyle Fraser Heart Center/Crawford Long Hospital, Emory University School of Medicine, 550 Peachtree Street, NE, Atlanta, GA , United States b Heart Foundation Research Center, Griffith University Gold Coast Campus, Southport, QLD, Australia Received 1 November 2004; received in revised form 18 February 2005; accepted 22 February 2005 Available online 1 April 2005 Time for primary review 21 days Abstract Objective: This study tested the hypothesis that brief cycles of iterative ischemia reperfusion at onset of reperfusion (termed bpostconditioningq, post-con) delays washout of intravascular adenosine and thereby increases endogenous adenosine receptor (AR) activation during the early moments of reperfusion (R). Methods: Isolated mouse hearts were subjected to 20 min global ischemia (I) and 30 min R with or without post-con (3 or 6 cycles of 10 s R&I). Intravascular purines in coronary effluent were analyzed by HPLC. To assess the functional role of endogenous AR activation in postcon, an open-chest rat model of myocardial infarction was employed. Rats were randomly divided into 11 groups: control, no intervention at R; post-con, three cycles of 10 s R followed by 10 s LCA re-occlusion immediately upon R. In the following interventions, drugs (or vehicle) were administered 5 min before R in the absence or presence (F) of post-con. Vehicle (DMSO b300 cl/kg); 8-SPT (non-selective AR antagonist, 10 mg/kg) F post-con; DPCPX (A 1A R antagonist, 0.1 mg/kg) F post-con; ZM (A 2A AR antagonist, 0.2 mg/kg) F postcon; MRS1523 (A 3 AR antagonist, 2 mg/kg) F post-con. Results: In isolated mouse hearts, post-con reduced diastolic pressure during both early (26F3* vs. 37F3 mmhg at 5 min) and late (22F3* vs. 34F3 mmhg at 30 min) R. Post-con also hastened the early recovery of contractile function (developed pressure 39F6* vs. 16F2 mmhg at 5 min R), although differences did not persist at 30 min R. Importantly, post-con was associated with reduced adenosine washout (58F5* vs. 155F16 nm/min/g) at 2 min R suggesting greater retention time of intravascular adenosine. In rats, post-con significantly attenuated infarct size compared to control (40 F 3% vs. 53 F 2%* in control), an effect that was unaltered by DPCPX (42 F 2%) but was abrogated by 8-SPT (50F2%), ZM (49F3%) or MRS1523 (52F1%) (Pb0.02). Conclusion: These data suggest that post-con involves endogenous activation of A 2A and A 3 but not A 1 AR subtypes. This activation may be linked to the delay in the washout of intravascular adenosine during the early minutes of R during which post-con is applied. D 2005 European Society of Cardiology. Published by Elsevier B.V. All rights reserved. Keywords: Postconditioning; Adenosine; Receptors; Infarction; Reperfusion 1. Introduction T Corresponding author. Tel.: ; fax: address: azatta@emory.edu (A.J. Zatta). The definitive treatment for myocardial ischemia is reperfusion. However, reperfusion has the potential to cause additional reversible and irreversible damage to the myocardium [1,2]. Reperfusion injury has been shown to be attenuated by pharmacological and mechanical interventions introduced at the onset of reflow. Accordingly, exerting control over the hydrodynamics of the early (i.e. first 30) minutes of reperfusion reduces post-ischemic injury in a canine model of acute coronary occlusion reperfusion [3,4]. These studies not only supported the concept that reperfusion is associated with additional injury, but also provided the /$ - see front matter D 2005 European Society of Cardiology. Published by Elsevier B.V. All rights reserved. doi: /j.cardiores

2 H. Kin et al. / Cardiovascular Research 67 (2005) scientific basis for the more recently reported phenomenon of bpostconditioningq (post-con) [5], which is a time-sensitive cardioprotective strategy induced by brief repetitive interruptions in blood flow applied immediately at the onset of reperfusion [5 8]. To date, cardioprotection by postconditioning has been reported by independent laboratories in several species (i.e. canine [5,7], rabbit [9], rat [6]), in isolated perfused hearts [10] and in in vivo and cell culture models [5,8]. In addition, it has been documented that postconditioning affords protection comparable to that of ischemic preconditioning (pre-con) [5,9,10], although there are mixed reports on whether these strategies are capable of invoking additive protection when combined [5,7,9,10]. Importantly, recent studies have provided direct evidence supporting a role for nitric oxide and the reperfusion injury salvage kinase (RISK) pathway [9,10]. However, there are no studies that have investigated potential upstream triggers of the signalling elements proposed. One likely candidate for a cardioprotective trigger stimulated by postconditioning is adenosine receptor activation. In the last 10 years, an abundance of studies have cited the cardioprotective effects of adenosine receptor activation during both ischemia and reperfusion [11 14]. Importantly, adenosine receptor activation during reperfusion appears to be of key importance during the first 30 min, exerting not only anti-neutrophil properties but also infarctsparing effects [11,15]. Despite mixed support for a role for A 1 ARs [16 19], ample evidence suggests that activation of A 2A AR and A 3 ARs mediates protection exerted during reperfusion [15,20 27]. Accordingly, we hypothesised that AR activation by endogenous adenosine, in particular activation of the A 2A AR and A 3 ARs, may be involved in mediating postconditioning cardioprotection. An additional aim of this study was to investigate whether the transient impediments to reflow during postconditioning alters intravascular adenosine concentrations during early reperfusion, thereby providing a rationale for adenosine receptor activation during the postconditioning phase. 2. Materials and methods Investigations conformed to the Guide for the Care and Use of Laboratory Animals published by the US National Institute of Health (NIH Publications No , revised 1996) In vivo rat surgical preparation Ninety-five male Sprague Dawley rats weighing g were initially anesthetized via intraperitoneal administration of sodium pentobarbital (40 mgd kg 1 ). A tracheotomy was performed, and the trachea was intubated with a cannula connected to a rodent ventilator (Harvard Rodent Ventilator Model 683, Holliston, MA). Rats were ventilated with O 2 -enriched room air at ~ breaths per minute, with tidal volumes set to 1.0 ml/100 mg body weight. Arterial po 2, pco 2 and ph were monitored (Stat Profile M, Nova Biomedical, Waltham, MA) during the experiment and values were maintained within a normal physiological range (ph ; pco 2, mmhg; and po 2, mmhg) by adjusting the ventilatory rate and tidal volume, or by intravenous administration of sodium bicarbonate. Body temperature was kept at a constant 37 8C by using an adjustable heating pad. The left carotid artery was cannulated with a 24-gauge angiocatheter connected to a fluid-filled pressure transducer to continuously monitor phasic and mean arterial pressure (MAP) and heart rate (HR) using the EMKA data acquisition and analysis system (EMKATechnologies Inc., Falls Church, VA). The right external jugular vein was cannulated for delivery of anesthesia and drug treatments. The chest was opened via a left thoracotomy through the fourth or fifth intercostal space, and the ribs were gently retracted to expose the heart. After pericardiotomy, a 6-0 proline (Ethicon, NJ) ligature was placed under the left coronary artery (LCA), and the ends of the suture were threaded through polyethylene tubing (PE-50) to form a snare for reversible LCA occlusion. Following surgical preparation (prior to LCA occlusion), a bolus dose of sodium heparin (100 U/kg) was administered to ensure heparinization throughout the experiment. The LCA was occluded by tightening the snare using a light weight hemostat. Ischemia was confirmed by transient decrease in blood pressure and cyanosis on the myocardial surface. Reperfusion was indicated by an epicardial hyperemic response and rapid disappearance of cyanosis Experimental protocol for in vivo studies All rats were subjected to 30 min LCA occlusion, followed by 3 h reperfusion. In those hearts that received drug treatment, antagonists (or vehicle) were given as a bolus dose 5 min prior to reperfusion to ensure proper receptor blockade. Rats were randomly assigned to one of eleven groups (n = 8 12 in each group, Fig. 1): 1: Control, no intervention either before or after LCA occlusion (n = 12); 2: Vehicle control, DMSO (b 300 Al/kg) alone (n =8); 3: Post-con, initiated immediately at the onset of reperfusion, 3 cycles of 10 s full reperfusion and 10 s reocclusion (total intervention time of 1 min) (n =10); 4: 8- SPT (10 mg/kg) alone (n = 9); 5: 8-SPT+Post-con, post-con stimulus invoked in the presence of antagonist 8-SPT (n =8); 6: A 1 AR antagonist DPCPX (0.1 mg/kg) alone (n =8); 7: DPCPX +Post-con (n =8); 8: A 2A AR antagonist ZM (0.2 mg/kg) alone (n =8); 9: ZM Postcon (n =8); 10: A 3 AR antagonist MRS1523 (2 mg/kg) alone (n =8) and 11: MRS1523+Post-con (n =8) Determination of infarct size Upon completion of the experiment, the LCA was reoccluded and the area at risk (AAR) was delineated by

3 126 H. Kin et al. / Cardiovascular Research 67 (2005) Fig. 1. Experimental protocol employed to test for a role for adenosine receptor activation in post-con triggered protection. injecting 1.0 ml of 20% Unisperse blue dye via the external jugular vein. The heart was rapidly excised and placed into 0.9% saline. An individual blinded to the protocol subsequently separated the LV from remaining tissue and thinly (~ 2 mm) cross-sectioned the LV before separating the AAR (unstained) from blue stained non-ischemic zone. The AAR was incubated for 10 min in a phosphate buffered 1% TTC solution at 37 8C, enabling a clear differentiation of necrotic tissue. Area of necrosis (AN) and AAR were determined by gravimetric analysis. AN was expressed as a percentage of the AAR (AN/AAR) Plasma creatine kinase (CK) activity Arterial blood (0.3 ml) was collected in EDTA tubes at baseline, end of ischemia and 3 h of reperfusion for CK quantitation to confirm morphologic injury (necrosis). Samples were centrifuged at 2500g for 10 min and stored at 708C until analyzed enzymatically (CK assay, Diagnostics Chemicals Limited, Oxford, CT). Plasma CK activity is expressed as IU/g protein Murine Langendorff buffer-perfused heart model Twenty-three adult male C57/Bl6 mice (8 12 weeks age, 30F0.5 g body weight, 130F4 mg blotted heart weight) were anesthetized with sodium pentobarbital (50 mgd kg 1 ). Hearts were isolated and prepared as described previously [28,29]. After 20 min stabilization hearts were switched to pacing (420 beatsd min 1 ) using silver electrodes (0.5 ms square pulses, 20% above threshold, typically 2 5 V) placed on the LV and a Grass S9 stimulator (Grass, Quincy, MA, USA). After a further 10 min equilibration, hearts were then subjected to 20 min global normothermic ischemia, followed by 30 min reperfusion. Pacing was terminated on initiation of ischemia and resumed after 2.5 min of reperfusion in all groups. The following post-con regimes were assessed: 1: Control, no post-con stimulus (n = 7); 2: Table 1 Hemodynamic variables during the course of the experiment Group Baseline 25 min ischemia 180 min reperfusion Heart rate MAP RPP Heart rate MAP RPP Heart rate MAP RPP Control 348F27 103F5 35F2 334F28 104F5 33F2 265F11 100F5 29F1 Post-con 313F13 106F4 32F2 321F25 100F2 30F4 266F10 101F4 27F1 8-SPT 309F12 114F4 33F2 291F6 107F7 31F1 268F8 101F3 27F1 Post-con+8-SPT 331F6 110F2 33F1 322F24 102F5 31F2 257F6 97F4 25F1 DPCPX 366F18 95F6 35F3 369F18 97F5 36F3 326F27 85F4 28F4 Post-con+DPCPX 339F21 95F7 33F4 339F15 92F5 31F2 320F19 89F6 28F2 ZM F8 100F6 35F3 350F17 92F6 33F4 298F17 114F8 35F4 Post-con+ZM F12 97F4 34F2 341F13 90F4 31F2 290F16 99F6 29F3 Vehicle 374F20 86F8 33F4 382F14 95F9 37F4 312F30 82F10 27F5 MRS F28 100F3 34F3 303F33 85F9 26F3 264F19 83F16 22F5 Post-con+MRS F26 110F5 37F3 333F29 104F3 35F4 296F20 107F4 32F3 MAP, mean-arterial pressure; RPP, rate-pressure product/1000. *P b0.05 vs. values in control group.

4 H. Kin et al. / Cardiovascular Research 67 (2005) Table 2 Infarct size data Group n Body weight (g) Heart weight (mg) AAR/LV (%) AN/AAR (%) Control F20 739F42 32F2 53F2 z Post-con F25 701F57 30F1 40F3T 8-SPT 9 277F5 624F29 33F2 52F3 z Post-con F10 610F17 32F2 50F2 z 8-SPT DPCPX 8 341F5 675F14 30F2 55F2 z Post-con F7 658F12 33F1 42F2T DPCPX ZM F9 563F16 33F2 51F2 z Post-con F9 573F19 32F3 49F3 z ZM Vehicle 8 372F8 725F17 31F3 53F3 z MRS F18 732F27 32F3 49F2 z Post-con+ MRS F8 683F21 31F1 52F1 z Values are expressed as meanfs.e.m. All treatment groups were assessed simultaneously using one-way ANOVA for individual parameters. T P b0.05 vs. values in control group. z P b0.05 vs. values in post-con group. Post-con, initiated immediately at the onset of reperfusion, 3 cycles of 10 s full reperfusion and 10 s re-occlusion (total intervention time of 1 min) (n =7); 3: Post-con, 6 cycles of 10 s full reperfusion and 10 s re-occlusion (total intervention time of 2 min) (n = 9) Analysis of venous purine efflux To determine if post-con-mediated protection involves modulation of extracellular purine concentrations, coronary perfusate flow was quantified using a cannulating Doppler flow-probe (Transonic Systems, Inc., Ithaca, NY, USA) in the aortic perfusion line and connected to a T206 flowmeter (Transonic Systems), and coronary venous effluent was sampled every minute for the first 5 min of reperfusion and then collected at 10, 20 and 30 min time points for the remainder of the reperfusion period. The samples were immediately frozen at 80 8C until analyzed by reversephase HPLC with PDA detection (Waters, NSW, Australia) for adenosine, inosine, hypoxanthine, xanthine and uric acid. Early (0 5 min) and total post-ischemic purine efflux values were calculated as the product of concentration reached in the reperfusion effluent (nmol/ml) reperfusion volume (ml/min/g wet weight) Chemicals The A 2A AR antagonist ZM was purchased from Tocris, Ellisville, MO. All other drugs were purchased from Sigma/RBI (Sigma, St. Louis, MO). 8-SPT was dissolved directly in 0.9% saline, whereas DPCPX, ZM24138 and MRS1523 were dissolved in b300 Al/kg DMSO. Vehicle Fig. 2. Effect of adenosine receptor antagonism on the infarct-sparing effects of post-con following 30 min occlusion and 120 min reperfusion. A) Effects of non-selective adenosine antagonist, 8-SPT (10 mg/kg) or A 1 AR inhibitor, DPCPX (0.1 mg/kg) in the absence or presence of post-con. B) Effects of A 2A AR inhibitor, ZM (0.2 mg/kg) or A 3 AR blocker, MRS1523 (2 mg/kg) in the absence or presence of post-con. All values are meansfs.e.m. *P b0.05 vs. control group; yp b0.05 vs. post-con group.

5 128 H. Kin et al. / Cardiovascular Research 67 (2005) studies showed that this concentration had no effect on hemodynamics or infarct size (Tables 1 and 2, Fig. 2). ZM is a highly selective A 2A AR antagonist with a 1000-fold A 2A AR:A 1 AR selectivity, 91-fold A 2A AR:A 2B AR selectivity and 500,000-fold A 2A AR:A 3 AR selectivity [30] and has very little, if any, activity against responses mediated by A 1 or A 3 ARs in vivo [31]. Furthermore, MRS1523 is a highly potent ligand at rat A 3 ARs with selectivity of 140 and 18-fold vs. A 1 and A 2A receptors, respectively [32]. In addition, 2 mg/kg of MRS1523 has been reported to be the minimum dose required to inhibit A 3 -dependent histamine release from 100 Ag/kg CI-IB- MECA in mouse [33] Statistical analysis All values are reported as mean F S.E.M. Data were analyzed using Statistica version 7 (StatSoft, Tulsa, OK). One-way analysis of variance (ANOVA) (i.e. AAR and infarct size data) and multi-way ANOVA with repeated measures (cardiodynamics and functional responses) were used as appropriate, with Student Newman Keuls posthoc test when significant effects were detected. In all tests P b 0.05 was considered indicative of statistical significance. MRS1523 (selective A 3 AR inhibitor, 2 mg/kg) alone had no effect on infarct size compared to control (Table 2, Fig. 2A and B). However, the infarct sparing effects of post-con were completely abolished by 8-SPT, ZM and MRS1523 (50F2%, 49F3% and 52F1%, respectively, P b0.02) (Fig. 2A and B). In contrast, the A 1 AR inhibitor DPCPX was unable to modulate the cardioprotective effects of post-con (42F2%). Plasma CK activity was comparable among the 11 groups at the end of baseline and ischemia (data not shown). In the control group, plasma CK activity after 180 min of reperfusion was shown to increase by 3-fold in comparison to baseline, a response which was markedly attenuated by post-con (25F6 vs. 46F3 IU/g protein in control) (Fig. 3). Administration of drug treatments alone did not alter the elevations in CK activity relative to the control group. Importantly, DPCPX did not modify CK activity in post-con group (Fig. 3A), whereas 8-SPT and MRS1523 was associated with greater CK activity at 180 min of reperfusion (Fig. 3A and B). Interestingly, despite inhibition of the infarct-sparing effects of post-con by ZM241385, the reduction in CK activity observed in the post-con group was only partially reversed by selective A 2A AR antagonism (36F3 vs. 3. Results 3.1. Hemodynamic data The hemodynamic data are summarized in Table 1. There were no significant differences observed between control and treatment groups for any variable [HR, MAP or rate-pressure product (RRP)] at any time point (baseline, end of ischemia, and 180 min of reperfusion). Thus, despite the fact that the postcon group resulted in a smaller infarct size, hemodynamic variables were not enhanced. Indeed, this apparent dissociation has been observed by others, notably Cohen et al. [34] in which infarct size was reduced by preconditioning by 80%, and regional wall motion was improved, but systemic hemodynamics were unaltered A 2A and A 3 AR (but not A 1 AR) antagonism abolishes the infarct-sparing effect of post-con The area placed at risk by LCA occlusion, expressed as a percent of left ventricular mass (AAR/LV), was comparable among the 11 groups (~ 30 33%) (Table 2). Infarct size, expressed as a percentage of area at risk (AN/AAR), was significantly reduced in the post-con group (40F3%) compared with the control rats (53F2%, P b0.01) (Fig. 2). Pretreatment with 8-SPT (non-selective AR antagonist, 10 mg/kg), DPCPX (selective A 1 AR inhibitor, 0.1 mg/kg), ZM (selective A 2A AR inhibitor, 0.2 mg/kg) or Fig. 3. Effect of adenosine receptor antagonism on the anti-necrotic effects of post-con, measured using CK activity assay. A) Effects of non-selective adenosine antagonist, 8-SPT (10 mg/kg) or A 1 AR inhibitor, DPCPX (0.1 mg/kg) in the absence or presence of post-con. B) Effects of A 2A AR inhibitor, ZM (0.2 mg/kg) or A 3 AR blocker, MRS1523 (2 mg/kg) in the absence or presence of post-con. All values are meansfs.e.m. *P b0.05 vs. control group; yp b0.05 vs. post-con group.

6 H. Kin et al. / Cardiovascular Research 67 (2005) F3 IU/g protein in post-con group, P b0.06) (Fig. 3B). Importantly, post-con in the presence of ZM was not significantly different to ZM alone ( P b0.4) or control hearts ( P b0.3). These data are largely consistent with the infarct size data Post-con delays washout of intravascular adenosine during early reperfusion and improves contractile recovery in buffer-perfused hearts In control hearts, post-ischemic contractile function failed to recover to pre-ischemic levels following 20 min ischemia and 30 min reperfusion. End-diastolic pressure (EDP) was significantly elevated (~ 30 mmhg), and LV developed pressure (LVDP) and rate-pressure product (RPP) recovered to approximately 45% of pre-ischemic levels. Coronary perfusate flow rate was not significantly Fig. 5. Effect of post-con on extracellular purine effluxes during reperfusion. A) Time course of adenosine washout during early reperfusion. B) Total extracellular adenosine efflux during reperfusion. C) Total extracellular purine efflux during reperfusion. All values are meansfs.e.m. *P b0.05 vs. control hearts. yp b0.05 vs. 3 cycle algorithm. Fig. 4. Effect of post-con on post-ischemic recovery of contractile function in buffer-perfused mouse hearts. A) Recovery of end-diastolic pressure, B) LV developed pressure and C) rate-pressure product. All values are meansfs.e.m. *P b0.05 vs. control hearts; yp b0.05 vs. 3 cycle algorithm. different to pre-ischemic values (data not shown). Interestingly, while the post-con algorithm of three cycles of 10 s reperfusion and 10 s ischemia was unable to modify contractile function in hearts, lengthening the duration of the algorithm to six cycles significantly improved post-ischemic systolic and diastolic function of hearts (Fig. 4). Relative to controls, EDP was significantly attenuated during early (5 min, 26 F 3 vs. 37 F 3 mmhg in control, P b0.05) and late (30 min, 22F3 vs. 34F3 mmhg in control, P b0.05) reperfusion (Fig. 4A). However, post-ischemic LVDP was significantly greater with 6 cycles of post-con during early (39F6 vs. 16F2 mmhg in control, P b 0.05) but not during late reperfusion (82 F 5 vs. 66 F 3 mmhg in control) (Fig. 4B). Interestingly, RPP was significantly greater in comparison to control hearts during 5, 10 and 20 min of reperfusion, but not 30 min of reperfusion (Fig. 4C). Importantly, improvement in early reperfusion contractile recovery

7 130 H. Kin et al. / Cardiovascular Research 67 (2005) observed in these post-con hearts was also associated with greater retention time of intravascular adenosine during early reperfusion (58F6 vs. 155F16 nm/min/g in control hearts, P b0.05) (Fig. 5A). At 1 min of reperfusion, the effluent adenosine concentration with the 10 s algorithm (3 or 6 cycles) of post-con was significantly lower than in controls with unbridled reperfusion. However at 2 min of reperfusion, 1 min after the three cycle algorithm had ended, the effluent adenosine concentration returned to control levels in these hearts. In contrast, the effluent adenosine concentration with the six cycle algorithm remained lower at 2 min of reperfusion (i.e. until the end of post-con stimuli), and was significantly higher than controls at 3 min of reperfusion. Post-con did not modify total intravascular adenosine (Fig. 5B) or purine (Fig. 5C) concentration, suggesting that post-con results in the retention of adenosine rather than altering the metabolism of adenosine or its metabolites. 4. Discussion Data from the present study suggest that cardioprotection invoked by postconditioning involves endogenous activation of adenosine receptors. Our results support a key role for both A 2A and A 3 AR subtypes, but not the A 1 AR, in the infarct-sparing effects of postconditioning. In addition, postconditioning results in a greater retention time of extracellular (i.e. intravascular) adenosine during the period of postconditioning in isolated hearts, which was also associated with improved recovery of contractile function. These novel data therefore demonstrate that the infarct sparing effects of postconditioning can be abrogated by blocking adenosine receptor activation, suggesting functional involvement of endogenously released adenosine. Moreover, delay in washout of intravascular adenosine with postconditioning may accordingly increase the presence and concentration of adenosine in the vascular (and interstitial) space. Indeed, the time frame of delay in immediate adenosine washout is consistent with the critical timing of the postconditioning intervention [6]. Thus, adenosine receptor activation may provide the trigger for activation of downstream signaling mechanisms involved in cardioprotection Postconditioning mediated cardioprotection involves endogenous activation of adenosine receptors during early reperfusion To assess the role of endogenous adenosine receptor activation in postconditioning mediated protection, we studied the infarct-sparing effects of postconditioning in the absence and presence of non-selective adenosine receptor antagonist, 8-SPT, and subtype specific adenosine receptor antagonists, DPCPX (A 1 AR inhibitor), ZM (A 2A AR inhibitor) and MRS1523 (A 3 AR inhibitor). The protective effects of postconditioning were abolished to a comparable degree by non-selective adenosine receptor antagonism, and by specific A 2A and A 3 AR blockade (Fig. 2B). In contrast, we found no role for activation of A 1 ARs in postconditioning mediated protection (Fig. 2A). These data suggest that endogenous activation of A 2A and A 3 ARs during reperfusion play a key role in triggering the infarctsparing effects of postconditioning. Similarly, CK activity closely matched changes in infarct size (Fig. 3A and B). A recent study by Boucher et al. demonstrated that administration of A 2A AR agonist, CGS21680, starting 5 min prior to reperfusion resulted in a significant reduction in infarct size, whereas the same regimen initiated 5 min after the onset of reperfusion was ineffective [27]. These studies are not only consistent with our finding that A 2A ARs are involved in triggering postconditioning mediated protection, but also reinforces the concept that timing of the intervention (i.e. early moments of reperfusion) is critical in initiating protection during reperfusion [6]. Moreover, previous studies suggest that the effects of adenosine during ischemia are mediated by A 1 and A 3 ARs [18,29,35 40], whereas effects during reperfusion are largely ascribed to A 2A and A 3 ARs [15,20 27], which is consistent with our findings that A 2A and A 3 ARs (but not A 1 ARs) play a significant role in postconditioning mediated protection. Indeed, despite some support for endogenous activation of A 1 ARs reducing post-ischemic contractile dysfunction [17,18], administration of A 1 AR agonists during reperfusion alone have been largely unsuccessful in limiting infarct size [16,19,41]. Interestingly, there is little evidence supporting cardioprotection by intrinsic activation of A 3 ARs during reperfusion. Indeed, the beneficial effects of A 3 ARs are reported to be dependent on exogenous activation [21,24,25,40], suggesting that interstitial adenosine concentrations during reperfusion are not sufficient to cause physiologically effective activation of A 3 ARs. This is of particular interest since the present study showed that activation of both A 2A and A 3 ARs by endogenous adenosine is involved in triggering the cardioprotection of postconditioning. More importantly, the results from the present study raise the possibility that endogenous adenosine concentrations are modulated by the postconditioning intervention, positioning adenosine as a stimulus or trigger of cardioprotection. Indeed, interstitial and extracellular adenosine concentrations have been previously reported to be ~ 5 AM and 2 AM, respectively, during the first 5 min of reperfusion in the rat [18,42], which is well in excess of the adenosine affinities for the A 1 and A 3 ARs of 10 nm vs. N1 AM, respectively, in rat [43]. However, while interstitial adenosine concentrations closely match intravascular levels under basal conditions [42], Lasley et al. demonstrated that washout of interstitial adenosine lags behind that of coronary venous adenosine levels following prolonged coronary artery occlusion [44]. These observations not only suggest that

8 H. Kin et al. / Cardiovascular Research 67 (2005) intravascular washout precedes interstitial washout within the first 5 min, but indicate that the rate of intravascular washout is relatively high, which emphasizes that adenosine concentrations in these spaces are rapidly changing during the crucial first minutes of reperfusion when postconditioning may be applied. In light of this, we assessed the hypothesis that postconditioning delays the washout of endogenous adenosine, ostensibly by transiently impeding blood flow, thereby increasing intravascular adenosine concentrations during the critical early moments of reperfusion. Accordingly, our study revealed that postconditioning decreased coronary effluent adenosine during the first minute of reperfusion previously found to be critical for postconditioning [6] (Fig. 5A), which presumably implies a greater retention of the purine in the intravascular space. Indeed, analysis of purine efflux collected at 1 min intervals during early reperfusion revealed that postconditioning increased retention time of extracellular adenosine expressed as reduced effluent concentrations during the first 2 min of reperfusion (i.e. during the postconditioning stimulus) and increased concentrations between 2 and 4 min when hearts experienced uncontrolled reflow (Fig. 5A). While simultaneous measurement of intravascular adenosine from effluent and interstitial adenosine using microdialysis are necessary to confirm this hypothesis, this remains a significant challenge since the rate and volume of interstitial samples is much smaller than that which can be collected over the same time frame. Alternatively, the transient ischemic periods could have increased adenosine production, thereby contributing to the intravascular concentration. However, total adenosine and purine efflux were unaltered by post-conditioning, which argues against this possibility (Fig. 5B and C). The delayed washout of intravascular adenosine during reperfusion was also associated with improved early recovery of contractile function in our buffer-perfused hearts (Fig. 4). Therefore, the present study demonstrates that postconditioning 1) can protect against myocardial stunning, and 2) protection can be demonstrated in cell-free (asanguinous) system as reported previously [8]. These data suggest that protection exerted by postconditioning is not solely dependent on the presence of inflammatory cells or other blood-borne elements, but may also modify injury processes directly in the extravascular compartment Cardioprotection triggered by A 2A and A 3 adenosine receptor activation The observation that blockade of either A 2A or A 3 ARs abrogated postconditioning mediated protection to a similar extent (i.e. infarct size comparable to control heart) suggests that A 2A and A 3 ARs may be acting on a common end effector or that an ball or noneq response by endogenous adenosine on the target effector is involved, whereby both receptors are required to be activated to manifest the response (i.e. one subtype alone cannot sufficiently modify signaling to produce the phenotypic change). Involvement of A 2A and A 3 AR activation triggered by accumulating adenosine during postconditioning is consistent with their ability to inhibit inflammatory-related processes [2,15,20 22,45,46], which are manifested as a reduction in coronary artery endothelial dysfunction, neutrophil accumulation, generation of reactive oxygen species and their lipid peroxide products in myocardium protected by postconditioning [5 7,20]. Indeed, previous studies have shown that decreases in infarct size exerted by A 2A and A 3 AR agonism involved reductions in neutrophil adherence and/ or infiltration [20,21,47] and that adenosinergic cardioprotection elicited by activation of A 2A ARs is also associated with reductions in ROS generation and endothelial adherence of neutrophils [20,46]. Additionally, inhibition of the coronary vascular endothelium by retained intravascular adenosine may attenuate the subsequent PMN-endothelial cell interactions, ROS generation and cytokine release that are observed during early reperfusion. Thus A 2A and A 3 AR activation may modify a range of inflammatory processes contributing not only to direct intravascular protection, but also the infarctsparing effects observed in the present study. However, as mentioned above, other mechanisms not involving a cell-mediated inflammatory response to reperfusion are likely engaged by postconditioning. Interestingly, activation of the A 2A and A 3 AR (but not A 1 AR) subtypes during reperfusion may stimulate downstream signaling elements leading to cardioprotection relevant to postconditioning. Indeed, it has been demonstrated that the beneficial effects of A 2A AR activation during early reperfusion are dependent on phosphatidylinositol 3-kinase (PI3-K) [27]. Similarly, activation of A 3 ARs has been linked to a reduction in apoptosis via the PI3-K pathway [48]. In contrast, however, there appears to be no role for the PI3-K/Akt pathway in A 1 mediated protection [49,50]. Thus, the very recent report by Yang et al. [51] that postconditioning was dependent on both endogenous adenosine receptor activation and PI3-K pathway, together with our findings that A 2A and A 3 AR activation trigger postconditioning, is consistent with the hypothesis that adenosine receptor activation by endogenous adenosine provides a proximal trigger that can activate reperfusion injury survival kinases. However, we did not investigate the signaling pathways involved in A 2A or A 3 receptor activation by endogenous adenosine during postconditioning in the present study Conclusions In conclusion, the current study demonstrates that endogenous activation of adenosine receptors, in particular the A 2A and A 3 AR subtype, is involved in postconditioningmediated cardioprotection in a rat model of myocardial

9 132 H. Kin et al. / Cardiovascular Research 67 (2005) infarction. In addition, our data suggest that postconditioning delays washout of intravascular adenosine during the critical first minutes of reperfusion, ostensibly increasing intravascular concentrations of endogenous adenosine during the postconditioning bstimulusq. We speculate that retention of intravascular adenosine may act either as a trigger of more distal cardioprotective signals, such as survival kinases, nitric oxide, and K ATP channels previously shown to be involved in postconditioning-mediated cardioprotection [9,10] or by directly modulating inflammatory-related processes. Collectively, these data imply that postconditioning results in an enhanced level of endogenous adenosine during the early moments of reperfusion, which is capable of activating adenosine A 2A and A 3 receptor subtypes to elicit protection against myocardial infarction. Acknowledgements This work was supported in part by NIH grant # HL and HL Amanda J. Zatta (nee Flood) was the recipient of a Postgraduate Research Scholarship from the National Heart Foundation of Australia and is a recipient of a Postdoctoral Fellowship from the American Heart Association. References [1] Piper HM, Garcia-Dorado D, Ovize M. A fresh look at reperfusion injury. Cardiovasc Res 1998;38: [2] Vinten-Johansen J. Involvement of neutrophils in the pathogenesis of lethal myocardial reperfusion injury. Cardiovasc Res 2004;61: [3] Vinten-Johansen J, Lefer DJ, Nakanishi K, Johnston WE, Brian CA, Cordell AR. Controlled coronary hydrodynamics at the time of reperfusion reduces postischemic injury. Coron Artery Dis 1992: [4] Sato H, Jordan JE, Zhao Z-Q, Sarvotham SS, Vinten-Johansen J. Gradual reperfusion reduces infarct size and endothelial injury but augments neutrophil accumulation. Ann Thorac Surg 1997;64: [5] Zhao ZQ, Corvera JS, Halkos ME, Kerendi F, Wang NP, Guyton RA, et al. Inhibition of myocardial injury by ischemic postconditioning during reperfusion: comparison with ischemic preconditioning. Am J Physiol Heart Circ Physiol 2003;285:H [6] Kin H, Zhao ZQ, Sun HY, Wang NP, Corvera JS, Halkos ME, et al. Postconditioning attenuates myocardial ischemia reperfusion injury by inhibiting events in the early minutes of reperfusion. Cardiovasc Res 2004;62: [7] Halkos ME, Kerendi F, Corvera JS, Wang NP, Kin H, Payne CS, et al. Myocardial protection with postconditioning is not enhanced by ischemic preconditioning. Ann Thorac Surg 2004;78: [8] Sun HY, Wang NP, Kerendi F, Halkos M, Kin H, Guyton RA, et al. Hypoxic postconditioning reduces cardiomyocyte loss by inhibiting the generation of reactive oxygen species and intracellular calcium overload. Am J Physiol Heart Circ Physiol 2004 [Electronic publication ahead of print]. [9] Yang XM, Proctor JB, Cui L, Krieg T, Downey JM, Cohen MV. Multiple, brief coronary occlusions during early reperfusion protect rabbit hearts by targeting cell signaling pathways. J Am Coll Cardiol 2004;44: [10] Tsang A, Hausenloy DJ, Mocanu MM, Yellon DM. Postconditioning: a form of bmodified reperfusionq protects the myocardium by activating the phosphatidylinositol 3-kinase-akt pathway. Circ Res 2004;95: [11] Lasley RD, Mentzer Jr RM. Protective effects of adenosine in the reversibly injured heart. Ann Thorac Surg 1995;60: [12] Vinten-Johansen J, Thourani VH, Ronson RS, Jordan JE, Zhao ZQ, Nakamura M, et al. Broad-spectrum cardioprotection with adenosine. Ann Thorac Surg 1999;68: [13] Linden J. Molecular approach to adenosine receptors: receptormediated mechanisms of tissue protection. Annu Rev Pharmacol Toxicol 2001;41: [14] Headrick JP, Hack B, Ashton KJ. Acute adenosinergic cardioprotection in ischemic-reperfused hearts. Am J Physiol Heart Circ Physiol 2003;285:H [15] Zhao Z-Q, McGee DS, Nakanishi K, Toombs CF, Johnston WE, Ashar MS, et al. Receptor-mediated cardioprotective effects of endogenous adenosine are exerted primarily during reperfusion after coronary occlusion in the rabbit. Circulation 1993;88: [16] Baxter GF, Hale SL, Miki T, Kloner RA, Cohen MV, Downey JM, et al. Adenosine A1 agonist at reperfusion trial (AART): results of a three-center, blinded, randomized, controlled experimental infarct study. Cardiovasc Drugs Ther 2000;14: [17] Finegan BA, Lopaschuk GD, Gandhi M, Clanachan AS. Inhibition of glycolysis and enhanced mechanical function of working rat hearts as a result of adenosine A 1 receptor stimulation during reperfusion following ischaemia. Br J Pharmacol 1996;118: [18] Peart J, Headrick JP. Intrinsic A 1 adenosine receptor activation during ischemia or reperfusion improves recovery in mouse hearts. Am J Physiol Heart Circ Physiol 2000;279:H [19] Zhao Z-Q, Nakanishi K, McGee DS, Tan P, Vinten-Johansen J. A 1 receptor mediated myocardial infarct size reduction by endogenous adenosine is exerted primarily during ischaemia. Cardiovasc Res 1994;28: [20] Jordan JE, Zhao ZQ, Sato H, Taft S, Vinten-Johansen J. Adenosine A 2 receptor activation attenuates reperfusion injury by inhibiting neutrophil accumulation, superoxide generation and coronary endothelial adherence. J Pharmacol Exp Ther 1997;280: [21] Jordan JE, Thourani VH, Auchampach JA, Robinson JA, Wang NP, Vinten-Johansen J. A 3 adenosine receptor activation attenuates neutrophil function and neutrophil-mediated reperfusion injury. Am J Physiol Heart Circ Physiol 1999;277:H [22] Lasley RD, Jahania MS, Mentzer Jr RM. Beneficial effects of adenosine A 2A agonist CGS in infarcted and stunned porcine myocardium. Am J Physiol Heart Circ Physiol 2001;280:H [23] Zhao ZQ, Budde JM, Morris C, Wang NP, Velez DA, Muraki S, et al. Adenosine attenuates reperfusion-induced apoptotic cell death by modulating expression of Bcl-2 and Bax proteins. J Mol Cell Cardiol 2001;33: [24] Maddock HL, Mocanu MM, Yellon DM. Adenosine A 3 receptor activation protects the myocardium from reperfusion/reoxygenation injury. Am J Physiol Heart Circ Physiol 2002;283:H [25] Auchampach JA, Ge ZD, Wan TC, Moore J, Gross GJ. A 3 adenosine receptor agonist IB-MECA reduces myocardial ischemia reperfusion injury in dogs. Am J Physiol Heart Circ Physiol 2003;285:H [26] Kis A, Baxter GF, Yellon DM. Limitation of myocardial reperfusion injury by AMP579, an adenosine A 1/2A receptor agonist: role of A 2A receptor and Erk1/2. Cardiovasc Drugs Ther 2003;17: [27] Boucher M, Pesant S, Falcao S, de Montigny C, Schampaert E, Cardinal R, et al. Post-ischemic cardioprotection by A 2A adenosine receptors: dependent of phosphatidylinositol 3-kinase pathway. J Cardiovasc Pharmacol 2004;43: [28] Flood AJ, Willems L, Headrick JP. Coronary function and adenosine receptor-mediated responses in ischemic-reperfused mouse heart. Cardiovasc Res 2002;55:

10 H. Kin et al. / Cardiovascular Research 67 (2005) [29] Peart J, Flood AJ, Linden J, Matherne GP, Headrick JP. Adenosinemediated cardioprotection in ischemic-reperfused mouse heart. J Cardiovasc Pharmacol 2002;39: [30] Poucher SM, Keddie JR, Singh P, Stoggall SM, Caulkett PW, Jones G, et al. The in vitro pharmacology of ZM , potent, non-xanthine A 2A selective adenosine receptor antagonist. Br J Pharmacol 1995;115: [31] Keddie JR, Poucher SM, Shaw GR, Brooks R, Collis MG. In vivo characterisation of ZM241385, a selective adenosine A 2A receptor antagonist. Eur J Pharmacol 1996;301: [32] Li AH, Moro S, Melman N, Ji XD, Jacobson KA. Structure activity relationships and molecular modeling of 3, 5-Diacyl-2,4-dialkylpyridine derivatives as selective A 3 adenosine receptor antagonists. J Med Chem 1998;41: [33] Peart JN, Gross GJ. Adenosine and opioid receptor-mediated cardioprotection in the rat: evidence for cross-talk between receptors. Am J Physiol Heart Circ Physiol 2003;285:H81 9. [34] Cohen MV, Liu GS, Downey JM. Preconditioning causes improved wall motion as well as smaller infarcts after transient coronary occlusion in rabbits. Circulation 1991;84: [35] Lasley RD, Mentzer Jr RM. Adenosine improves recovery of postischemic myocardial function via an adenosine A1 receptor mechanism. Am J Physiol Heart Circ Physiol 1992;263:H [36] Tracey WR, Magee W, Masamune H, Kennedy SP, Knight DR, Buchholz RA, et al. Selective adenosine A 3 receptor stimulation reduces ischemic myocardial injury in the rabbit heart. Cardiovasc Res 1997;33: [37] Hill RJ, Oleynek JJ, Magee W, Knight DR, Tracey WR. Relative importance of adenosine A 1 and A 3 receptors in mediating physiological or pharmacological protection from ischemic myocardial injury in the rabbit heart. J Mol Cell Cardiol 1998;30: [38] Liem DA, van den Doel MA, de Zeeuw S, Verdouw PD, Duncker DJ. Role of adenosine in ischemic preconditioning in rats depends critically on the duration of the stimulus and involves both A 1 and A 3 receptors. Cardiovasc Res 2001;51: [39] Peart J, Willems L, Headrick JP. Receptor and non-receptor-dependent mechanisms of cardioprotection with adenosine. Am J Physiol Heart Circ Physiol 2003;284:H [40] Tracey WR, Magee WP, Oleynek JJ, Hill RJ, Smith AH, Flynn DM, et al. Novel N6-substituted adenosine 5V-N-methyluronamides with high selectivity for human adenosine A 3 receptors reduce ischemic myocardial injury. Am J Physiol Heart Circ Physiol 2003;285: H [41] Xu Z, Downey JM, Cohen MV. AMP579 reduces contracture and limits infarction in rabbit heart by activating adenosine A 2 receptors. J Cardiovasc Pharmacol 2001;38: [42] Harrison GJ, Willis RJ, Headrick JP. Extracellular adenosine levels and cellular energy metabolism in ischemically preconditioned rat heart. Cardiovasc Res 1998;40: [43] Ganote CE, Armstrong SC. Adenosine and preconditioning in the rat heart. Cardiovasc Res 2000;45: [44] Lasley RD, Hegge JO, Noble MA, Mentzer Jr RM. Comparison of interstitial fluid and coronary venous adenosine levels in in vivo porcine myocardium. J Mol Cell Cardiol 1998;30: [45] Cronstein BN. Adenosine, an endogenous anti-inflammatory agent. J Appl Physiol 1994;76:5 13. [46] Zhao Z-Q, Sato H, Williams MW. Adenosine A 2 -receptor activation inhibits neutrophil-mediated injury to coronary endothelium. Am J Physiol Heart Circ Physiol 2004;271:H [47] Cronstein BN, Rosenstein ED, Kramer SB, Weissmann G, Hirschhorn R. Adenosine: a physiologic modulator of superoxide anion generation by human neutrophils. Adenosine acts as an A 2 receptor on human neutrophils. J Immunol 1985;135: [48] Gao Z, Li BS, Day YJ, Linden J. A 3 adenosine receptor activation triggers phosphorylation of protein kinase B and protects rat basophilic leukemia 2H3 mast cells from apoptosis. Mol Pharmacol 2001;59: [49] Qin Q, Downey JM, Cohen MV. Acetylcholine but not adenosine triggers preconditioning through PI3-kinase and a tyrosine kinase. Am J Physiol Heart Circ Physiol 2003;284:H [50] Regan SE, Broad M, Byford AM, Lankford AR, Cerniway RJ, Mayo MW, et al. A 1 adenosine receptor overexpression attenuates ischemia reperfusion-induced apoptosis and caspase 3 activity. Am J Physiol Heart Circ Physiol 2003;284:H [51] Yang XM, Philipp S, Downey JM, Cohen MV. Postconditioning s protection is not dependent on circulating blood factors or cells but involves adenosine receptors and requires PI3-kinase and guanylyl cyclase activation. Basic Res Cardiol 2005;100:57 63.